After 24 straight hours of development, a half-dozen finalist teams — out of more than 50 — presented their opioid apps on December 7th at the finale of the U.S. Health & Human Services (HHS) Opioid Code-a-Thon. Kevin Merritt, founder and chief executive officer of Socrata, spoke as judges made their final decisions on which teams would win the three $10,000 prizes, highlighting the city of Cincinnati’s data set — the sole local government developed data set of the more than 70 data sets HHS prepared for the event.
Dayton, Ohio, is currently the most severe epicenter of opioid morbidity, Merritt noted, but Cincinnati, just 45 minutes away, had also found itself hard hit. “The city turned to data with amazing results,” he said. Cincinnati’s “heroin dashboard” has been a chief tool in the city’s ability to fight back the opioid epidemic. One change, Merritt noted, is the city’s ambulances are re-positioned in order to save lives much faster.
HHS in preparing the massive data trove — including previously unavailable data sets — is setting the national stage to ignite data-driven opioid apps like Cincinnati’s in order to best the opioid threat on three fronts — prevention, usage and treatment.
This is the first time we have done something of this scale,” said Mona Siddiqui, HHS chief data officer said.
Bruce Greenstein, HHS Chief Technology Officer, said that the University of Louisiana at Lafayette will be hosting the next opioid code-a-thon using the HHS opioid data trove on April 13-14th, 2018. The teams that created the code-a-thon’s winning opioid apps, as well as others, have the opportunity to apply for HHS funding to get their innovations to market and into the hands of local and state governments that need the insights the most.
Opioid Prevention Track Winner Focused on Drug Take Back Mapping
The team from Visionist, Inc., of Columbia, Maryland, including presenter Taylor Corbett, used Centers for Disease Control (CDC) data along with data from Walgreens, CVS and other private drug drop companies, to reduce those at-risk for opioid abuse from getting the drugs in the first place.
More than 70 percent of opioid users start by getting opioid medications and substances from their families, said Corbett.
While the tool can be scaled nationally, in less than 24 hours the team created a five-state regional platform that displays a risk index for potential opioid abuse. The “Takeback America” tool provides visualizations of hard hit counties and the availability of drug take-back programs.
By revealing which hard hit areas have low availability of take back programs, public health decision makers can focus resources where they are needed to increase participation and availability.
Also presenting was the finalist team Protecting the Next Generation of Ryan Haight, who died from purchasing Vicadin online and became the namesake for 2008 legislation that “prohibits the delivery, distribution or dispensing of a controlled substance that is a prescription drug over the Internet without a valid prescription,” according to Govtrack.us.
Tim Mackey of University of California San Diego School of Medicine, speaking for the team, said they applied machine learning to find opioids for sale on social media. Their tool analyzed 120 tweets made between November 15th and December 5th, finding 40 online pharmacies — and their locations — willing to sell opioid drugs without a prescription. The team used U.S. Food and Drug Administration, Drug Enforcement Agency, mortality and census data sets in ArcGIS, automating epidemiological surveillance in real-time. Law enforcement can query date ranges and keywords to detect, analyze and score the online pharmacies, Mackey said.
A third finalist team in opioid apps for prevention was from the University of North Carolina. This team created a patient-facing tool using Carolina Data Warehouse for Health data called the tHOR app. The goal of the tool is to provide patients with a more robust opioid risk assessment.
Opioid Usage Track Winner Builds Instrumentation Panel for Physicians Prescribing Opioids
Team OPAT (Opioid Prescriber Awareness Tool) took the stage at HHS headquarters asking everyone to step into the moment when a physician is with a patient.
“As a society, we have changed the rules for you as a physician,” team OPAT explained, noting a litany of training and liability issues that comprise physician realities in treating patient pain.
The team incorporated data from the Centers for Medicare & Medicaid Services to build a one-stop shop for physicians to turn to when pain is in question, and a red flag goes up.
- Compares physicians to other prescribing doctors in their area
- Gives information about specialists in the area, and their tendencies to prescribe opioids
- Offers referrals for behavioral health resources nearby
- Links directly to state opioid registry data
The purpose is to enable physicians to offer a plan that minimizes a patient’s potential opioid usage to the greatest extent possible.
Opioid Treatment Track Winner Forecasts Opioid Overdose Spikes for Emergency Response
The team Origami Innovations out of Yale University comes from hard-hit Connecticut. Their opioid app was born from a shortage of naloxone on June 23, 2016 that led to three deaths at Yale New Haven Hospital. An overdose outbreak brought 12 people to the hospital, all found in a one-block radius.
“If they could have predicted an outbreak, they could have prepared and saved lives,” said team Origami Innovations.
So, they created their application for first responders and hospitals — the agents that most need to know when they need more beds and more naloxone in order to reduce morbidity and increase access to treatment.
The application forecasts spikes in opioid overdoses in situations where limited data is available, such as in rural communities.
By using predictive analytics, and incorporating geospatial visuals, when a spike hits one county, the application can advise nearby counties in activating emergency response. The application applies a Poisson distribution.
The team used Accidental Drug-Related Deaths Data for Connecticut from HHS, which offers data on both space and time levels. The team says the platform is reliable and can “save lives as soon as tomorrow” because it uses real-time data as it comes in.
Users can select their county, and switch views from heroin and fentanyl, and the tool applies confidence intervals for the expected fallout in nearby areas. The application can be scaled for use in other states, and emergency responders can access the live application through mobile devices.
Also presenting was finalist Text to Connect, which used Substance Abuse and Mental Health Services Administration data for their opioid app. The text-based tool is designed to support families in rural areas when addicted people encounter road blocks to recovery, such as when they can’t get a bed in a treatment facility. Text “connect” to 21333 for quick access to care and to access a six-month behavioral education program. “Empowerment starts with good content,” said the team.
Finally, Pivot Health presented a solution to re-brand naloxone kits and make them real for drug users. According to Jordan Lahey, 80 percent of people do opioid drugs by themselves, and naloxone adoption is “awkward.” He envisions drug courts providing the opioid app his team developed, which focuses on buddy usage that earns users medication-assisted treatment credits and rewards for needed things like rent money and bus tickets. Pivot Health seeks to both provide a path to recovery for opioid users, and through reporting to local governments, a better way to curate workflows for existing government programs.
— HHS IDEA Lab (@HHSIDEALab) December 7, 2017
Watch the code-a-thon concept presentations and awards on the HHS Facebook page: